FINAL REPORT Period covered: October 1, 2001 - January 31, 2664 The Effects ofHighways on Fragmentation of Small Mammal Populations and Modifications of Crossing Structures to Mitigate Such Impacts Prepared by: Kerry R. Foresman Profes s or of Biology Division of Biological Sciences University of Montana Missoula, MT 59812 Prepared for: MONTANA DEPARTMENT OF TRANSPORTATION RESEARCH SECTION 2701 PROSPECT AVENUE HELENA, MT 59620 March 2004 Technical Report Doc u menial ton Page i. nit* it*. FHWA I 'UT-04*0Q5 t bitiTiniii Acctuitt rt< 3. n*t (plant ■ ClIaEa* No 4. Till*. . \i »lt1 hit. The HfTectsorHlizh«;iy^ on K ragmen tat Jon of Small Mammal Populations and Modifications of Crossing Structures to Mitigate Such Impacts 3 report □ ■!■ March 2004 Ptrf srmlnq i.i'iul; i' to t. Autiitrf*) Kerry R. Foresman r Paffanninj Organisation n«p*it tin. A- P*rt4t1tllrt| Crg>nl'MlDrt NaaMa laid A4#*a*a Division of Biological Sciences The University of Montana Missoula, MT 59812 i«* warh u*n no. I I. Ctltilil 6. till I UM M27334 ; MDT Project £Si6i 17. t|#ni#Mrtf *[*r him ir( - 111*11 Research Section Montana Department of Transportation 2701 Prospect Avenue PO Box 201001 Helena MT 59620-1001 l »♦ T|p* »| nap ft rt ft Pa rift I Civafftd Final Report: October 1.2001 - February 29. 2004 5401 i5. *up»M->Miir r urn Research performed in cooperation with the Montana Department of Transportation and the U. S~ Department of Transportation, Federal Highway Administration. This report can be found at + v. ■•■• w ndt.state m i.t i re : -uch docs resea re h_projanima l_u*epha sell final_reportpdf I. Attlrttt High vj ay 93 south from Mssoula to the Idaho border is currently being widened from 2 lanes to 4* Thij construction project has incorporated the use of small diameter, drainage culverts fitted with animal shelving in an attempt to mitigate for habitat fragmentation and loss of animal population continuity* The present research was onducted to determine the effectiveness of such shelving and to modify the design, where necessary, so that the widest variety of spedes could be served* Three 1*2 m diameter steel corrugated culverts were fitted with shelving and 3 additional culverts adjacent to hese P without shelving, were chosen to serve as controls* Over the period November 1, 2001 to January 31. 2004 emote sensing 35 mm cameras placed in each culvert recorded animal activity* Water levels within culverts was measured on a weekly basis and environmental variables (temperature* light and humidity levels) were recorded djacentto these culverts at 10 minute intervals* Relative abundance of small mammal species adjacent to the ntrancesof each culvert were determined through use of live -trap transects. O/er 4,500 photographs and 8* 100 events of animal activity (movement past detectors) were recorded (documenting culvert use by 14 different mammal species* In general the amount of use by a species reflected its elative abundance adjacent to the culvert- When water was present in the control culverts* animal use was greatly mited, if not completely prohibited. Using photographs which illustrated the behavior of animals while using shelving, hi edifications were made to increase the effectiveness of the shelf surf ace and entrance ramp. Following modifications, helving allowed for continued use of culverts even when water was present. Fourteen species of mammal were observed to use the shelves* Awmal movement patterns were highly seasonal reflecting responses to cold emperatures, standing water levels which prohibited access to culverts (and shelving ramps)/ and species behavioral differences* Vegetative cover at culvert entrances greatly influenced culvert use: protective cover was required in order or most species to move to the culverts. Meadow voles did not use culverts until a protected tube was provided which bould be accessed without having to leave the surrounding vegetation* There was no strong correlation between traffic 'olumes and animal activity- However most species at these locations were nocturnal and thus their movements occurred at night when traffic was light. Animal use of larger culverts with, and without animal shelves was also studied* Recommendations for shelf use land for modifications of other culvert t/pes is provided. 17. K*r words Culvert animal t small mammals helves, wildlife crossing structurei it.oiitHtaiun atut-iftii Unrestricted. This document is available through the National Technical Information Service, Springfield. VA 21161. lfl + Sal Unci |R> CLatllf- ssitied ■ ii It* Ill Unci iR) ciittif* \et ihi* ssified 21 NotfttPi :45 ; : . - 1 i » DISCLAIMER STATEMENT This document is disseminated under the sponsorship of the Montana Department of Transportation and the United States Department of Transportation in the interest of information exchange. The State of Montana and the United States Government assume no liability of its contents or use thereof The contents of this report reflect the views of the authors* uho are responsible for the facts and accuracy of the data presented herein* The contents do not necessarily reflect the official policies of the Montana Department of Transportation or the United States Department of Transportation The State of Montana and the United States Government do not endorse products of manufacturers. Trademarks or manufacturers 1 names appear herein only because they are considered essential to the object of this document This report does not constitute a standard, specification* or regulation, ALTERNATIVE FORMAT STATEMENT The Montana Department of Transportation attempts to provide reasonable accommodations for any known disability that may interfere with a person participating in any service, program, or activity of the Department. Alternative accessible formats of this document will be provided upon request* For further information* call (406)444-7693 or TTV (406)444-7696, ACKNOWLEDGMENTS I if ould first and foremost like to thank Jeremy Moran, my research technician, for all of the lab and field work that went in to making this research not only possible* but also so productive* He was instrumental in helping me solve the daily problems that arise in such research and in helping me juggle the myriad of projects that we focused upon as we continuously expanded this work. I would also like to thank Pat Basting, Dwane Kailey. and Loran £, Frazier, all within the Montana Department of Transportation (MDT) r for their help on various phases of this research. Pat in particular, took a keen interest in what we were doing and helped us address many issues that arose requiring MDT support. Gordon L Larson of Peccia & Associates helped in our earliest discussions of wildlife structures and was very receptive in helping us implement our ideas in on-going as well as future highway projects* Cory M* Qaussen of Roscoe Steel & Culvert Co* also worked closely with us to incorporate the modifications which we determined were necessary in the development of a very effective small animal culvert sheff* Numerous undergraduate students helped in many phases of this research. Among them were; Traynor Biasiolli, Erin Bohman* Kimberly Siring, Cory Engkjer, Randy Fenster, Jessica Gudgel, Dan Haskell Shannon Hickey, Ann Johnson, Michael Maraghy, Lauren Martin* Jessica Meredith, Jeremy Mizel, Dylan Mrkich, Timothy Pit;.- Nathan Schwab/ Ray Selbe, Brian Silva, Kate Soetaert, Dustin Thompson, Lindsay Whitehurst* and Karla Zarbock. All of these individuals made this research possible and I am indebted to them* Conversion Chart Metric Enflish 1 meter (m) 39*37 inches (in) 1 centimeter (cm) 0.3937 Inches 2.54 cm 1 inch TABLE OF CONTENTS Technical Report Documentation P.ge ... ii Disclaimer* Alternative format. Acknowledgements _.. ii i Metric to English Conversion Chart _..iii Table of Contents iv List of Figures and Tables . v Introduction _ „ _ _ 1 Materials and Methods- _ _ _ .. 2 Objective # 1: Modify shelves to enhance animal use 2 Objective #2: Development and test of effectiveness of small mammal tubes to accommodate selective species . 7 Objective **3: Determination of the importance of vegetative cover at/ and adjacent to* culvert entrances 8 R e su ks _ _ _ _ _ _ « _ _ - _ . . 9 Animal use of culverts 9 Effectiveness of small mammal shelf tubes 16 Animal activity at additional culverts . 18 Importance of vegetative cover at culvert entrances 19 Traffic volumes — _ . 22 Animal presence adjacent to main culverts 23 Conclusions « _ _ « 24 Recommendations « 27 Im p I e m e n tati on « _ . 27 References cited _ „ _ 28 Appendix I: Main culverts in this study and vegetative cover at their e ntr a n ce _ _ _ . 33 Appendix II: Additional culverts added during this study . 36 Appendix III: Final design of animal shelves 38 LIST OF FIGURES AND TABLES Figure 1* Original culverts, (a) and (b)« and 25 mm diamond shelving (c) 2 Figure 2. Location of culverts in which cameras >» ere installed _ _ _ 4-5 Table 1* Specifications of all culverts studied . 6 Figure 3* Remotely activated cameras mounted in the culverts: 35 mm (a), and video W— — — — — -- — 7 Figure 4. Small mammal live-trapping design _ _ _ 7 Figure 5* Prototype design of vole tubes and entrance ramps: (a) gutter tubes* (b) connected tubes suspended from shelf/ and (c) entrance ramp to vole tube 8 Figure 6* Sooted-tracking plate (a) placed in center section of vole tube (b) 8 Figure 7* Vegetation quadrants sampled at entrances to each main culvert 9 Table 2* Summary of animal activity in culverts from October 2001 through January 2004 _ _ _ _ _ _ _ _ * . 9 Table 3* Animal use of culverts as documented by photographs (October 2001 through January 2004....} « » 10 Figure 8* Seasonal activity pattern of animal use for all culverts .. 11 Figure 9* Animal activity in the culverts as a function of standing water levels ..11 Figure 10* Activity as a function of environmental temperatures in February (a) and July (b) 2002, and February (c) and July (d) 2003 _ _ _„.... ., 12-13 Figure 11. Raccoon wading through Gravel Pit Control culvert _ 14 Figure 12. Examples of 2 species using original shelves: Short-tailed weasel (a) and deer mouse (b) using the frame rather than the floor mesh of the original shelves- ~ « « « _. 14 Figure 13* Examples of 2 species using solid plastic surface of the shelving: Deer mouse (a)/ and painted turtle (b)_ 15 Figure 14* Examples of 2 species using modified shelves: Mountain cottontail (a) and porcupine (b) _ _ _» _• 15 Figure 15* Entrance ramp at Carleton Creek Experimental positioned to side of shelving and cu I v e rt ~ _ 15 Figure 16* Photos of nen shelf and entrance: (a) frontal vie>i with entrance ramp entering from side, (b) side view, and (c) top view showing width of entrance ramp and shelf *_ .. 16 Figure 17. Appearance of entrance funnel entering from side (plywood model)- 16 Figure 18. Animal use of vole tubes illustrated by sooted-tracking plates: (top) meadotr vole and deer mouse tracks and (bottom) short-tailed weasel tracks and drag mark of prey_ _ « _ _ 17 Figure 19. Cross section (a) of final shelf design illustrating vole tube incorporated in to the shelf frame and access funnel (b) built in to the shelf s entrance ramp .. 17 Figure 20* Red fox (a), and coyote (b) using 190 Large culvert irrth a silted floor surface _ 18 Figure 21* White tailed deer using 190 Large (a), and Bass Creek Large (b) culverts 18 Figure 22* Modified animal shelf placed in Bass Creek Large culvert . 19 Figure 23. Vegetative cover at the entrances to each main culvert 20 Figure 24. Height of vegetation at the entrances to each main culvert _ 21 Figure 25* Relationship between vegetative cover adjacentto entrances of the main culverts and animal activity (as indexed by total number of photographs recorded) _ _ - _ _„ «_. 22 Figure 26. Weekly traffic counts along Highway 93 between Carleton Creek road and Lolo, Montana- _ __ «_ _ __ __ . 22 Figure 27* Summary of small mammal live-trapping sessions over the period March 2001 to September 2003-.-. _ _ __ 24 Figure 28* Carleton Creek Experimental Culvert _„ __ 34 Figure 29. Carleton Creek Control Culvert « _.. 34 Figure 30* Gravel Pit Experimental Culvert -- ~ 34 Figure 31* Gravel Pit Control Culvert _ _„ _„ «_ _♦. 35 Figure 32* Maclay Flats Experimental Culvert. « .. 35 Figure 33* Additional culverts along Highway 93 South: (a) Gravel Pit Large, and (b) Bass Creek Large culverts 37 Figure 34* Additional culvert added along Highway 203: Florence culvert 37 Figure 35* Additional culverts added along Interstate 90: (a) 190 Large, (b) 190 Small, and (c) Double culverts . 37 Figure 36* Final design of animal shelves .„ » 39 Table 4* Small mammal species captured adjacent to culverts 23 VI INTRODUCTION In the 19 30' s people began to document the threat which highways posed to animals {1,2). As highways have widened and vehicle speeds have increased over the past 70 years, an increasing number of wild animal (and human) mortalities have been observed {3,4,5,6,?)* Data suggests that many larger species, such as wolves [Cants lupus, (8)] and grizzly bears [Ursus arccos (9)], prefer not to cross these open expanses while others such as deer {0 do cot feus sp*) and elk [Cervus efaphus). may be attracted to them (for salt or the abundance of green vegetation along the roadside) with dire consequences. Though less research has been conducted on the fragmentation effects that highw ays may present to small mammal populations, if hat data does exist suggests that such effects may be significant* Smaller mammals are actually impacted in 2 primary ways. First the wide expanses of a 4-lane highway (and shoulders) provide a formidable barrier for species that are primarily prey for mammalian carnivores and raptors* species such as shrews (Order Insectivora) and rodents (Order Rodentia). These species are very vulnerable rf they attempt to cross a highway, given the lack of protective cover, Indeed early trapping studies by Oxley et al. (10) found that dispersal of small mammals across a divided highway (a distance of 90+ m) was significantly reduced* consistent to the barrier effect posed by open bodies of water {ll t 12). They concluded that such a highway "-..may be as effective a barrier to the dispersal of small forest mammals as a body of fresh water twice as wide" (70:57)* A majority of these species have indeed evolved behavioral patterns that cause them to avoid areas devoid of vegetative cover thereby reducing their susceptibility to predation and* as a result, their populations are readily fragmented by highway construction. Secondly species that do attempt to cross these openings [e*g* the predators which behaviorally do not perceive openings as threatening such as striped skunks {Mephitis mephitis), w easels (Mustek sp*), raccoons (Procyon htor). red fo::es (Vulpes vufpes)> coyotes (Cants fa trans)] are often killed in this attempt because the distance is great and vehicles are moving so quickly* Such effects on these species are greatly magnified when highways bisect unique habitats such as wetland communities or forested areas which have historically served as wildlife corridors, areas which are often relatively small in overall size* Beginning in the 1980's and continuing today, studies have focused on the use of wildlife underpasses and culverts by animals, a majority of such research being conducted in Europe {13,14,15,16,27,18), More recently, studies in the United States have begun, as is evidenced by research presented at 5 international conferences on wildlife ecology and transportation {19,20,21,22,23) and by the recent compendium entitled Road Ecology published in 2003 (24)* At the moment, research falls into 1 of 2 areas, that of (l) identifying how animals perceive highways and railways as potential barriers (e.g*. 25,26, 27,2S) t and (2) mitigating for such impacts (e.g., 29,30,31,32,33)* The latter efforts involve development of safe passageways over or under such barriers to allow unimpeded movement for a variety of wildlife species* Importantly, any mitigation procedures which are implemented need to be validated* Many studies have begun to assess the effectiveness of such procedures {32 ,34,35 ,36,37 ,38,39) the results of which will allow for their further modification in the future. Assessing the state of our knowledge to date [an excellent review is presented in (24)]. it becomes abundantly clear that further documentation of these fragmentation effects are needed since data from many of these earlier studies if ere anecdotal or the studies themselves were limited in scope and experimental design* Additionally, though culverts and other such devices to allow movement of animals under or over highways have been shown to be used* to date little experimental rigor has been applied in most studies and few detailed conclusions have been drawn about proper design of such devices* With this background as a foundation, we conducted preliminary research in 2001 on a new highway project in west central Montana* The widening of Highway 93 from Lolo, Montana to the Idaho border began in several phases* one of which was identified as the "Lolo South Project" encompassing the section between Lolo and Florence. During this expansion Montana Department of Transportation (MDT) modified several culverts placed along wetland zones to include a ramp and w a Ik » ay- shelf. The hope was that small animals might use such platforms to move through the culverts even if hen water was present. Our preliminary research (40) suggested that certain species (e.g. raccoons, striped skunks* short-tailed weasels [Nustefo ermine*), deer mice {Peromyscus mantcufotus)] routinely used these culverts to traverse the highway while others, such as the meadow vole (Ntcrotos pennsylvanicos), which is the most abundant species living adjacent to the culverts, did not use them* In addition! species which used the culverts are influenced by water in the culvert, opting to use the shelf when water was present* Additionally! live-trapping data suggested that vegetation within the barrow pits significantly affected species distributions and that areas at the entrances of culverts which are devoid of vegetation limited animal movement to the culverts* From these preliminary results 3 objectives emerged which framed the following research; (1) Modify shelf structures to enhance small mammal use during wet periods. (2) Modify shelf structures to accommodate vole species, (3) Analyze the importance of a continuum of vegetative cover from the barrow pit to the culvert entrance. MATERIALS AND METHODS Objective #1: Modify shelves to enhance animal use Montana Department of Transportation installed animal shelves in 3 1.2 m diameter steel drainage culverts during construction of the Lolo South segment of Highway 93 (Fig* 1; Appendix I). We paired each of these culverts with an adjacent culvert not containing shelving, which then served as an experimental control [designated as Carleton Creek Experimental and Control (CCE and CCC). Gravel Pit Experimental and Control (GPE and GPC), and Maclay Flat Experimental and Control (MCE and MCC) (Fig* 2); MCC was subsequently found to be continually flooded and was removed from further study]* These remaining 5 culverts served as the primary focus of this research and will be referred to as the "main culverts* throughout this report* As this research progressed the number of culverts monitored on a daily basis increased to 12 (including culverts on 190 and Highway 203)* spanning a wide range of sizes and configurations (Fig* 2, Table 1, Appendix II)* Two of these additional culverts [Gravel Pit Large (GPL) and Bass Creek Large (BCL)) were also eventually retrofitted with modified animal shelving* Our preliminary research (40) suggested that modifications would be needed in the shelf design, in particular the floor surface and the entrance ramp. To address the floor surface, 2 diamond mesh sizes (25 mm and 6 mm)/ and one solid surface (heavy gauge polypropylene truck bed liner)/ were subsequently tested* Entrance ramps were repositioned upwards and a prototype of an entrance ramp design fabricated out of wood and e::truded sheet metal was tested on all shelves* m MmM Figure 1* Oiginal culverts, (a) and (b) and 25mm diamond shelving (c). Remote 35 mm cameras (Model TM 550.- TrailMaster'.. Goodson and Associates* Inc. Lenexa, KS) Mere installed to the roof of each culvert approximately 10 m inside the east entrance (Fig* 3 a)* These cameras ir ere activated by dual sensors an infrared detector which responded to heat and a microwave detector which responded to motion* thus any warm-blooded animal passing in front of the cameras was photographed* Each camera was capable of taking 36 color slides (Kodak Elite Chrome< ASA 200): if all 36 exposures were expended prior to weekly checks, the detector continued to document animal events and these w ere recorded. Though species determinations could not be made for these additional events* they did provide valuable information on overall use and activity patterns* In 2003, a remotely activated video camera (Model TM 700v, TrailMaster'', Goodson and Associates* Inc* Lenexdi KS) was installed in one culvert and periodically moved to other locations to record animal behavior within culverts (Fig* 3 b)* Environmental data loggers (Hobo' H08-004-02 or Hobo M Pro Series H08-032-08, Onset Computer Corporation* Pocasset. MA) Mere placed adjacent to the culverts, to record ambient temperature, light and relative humidity levels* These measurements if ere recorded at 10 minute intervals throughout the 2*5 year study. A passive infrared vehicle counter (Traffic Tally 3', Diamond Traffic Products, Oak ridge. OR) was installed at the GPE culvert (which lies approximately mid-way between the main culvert pairs) in June 2002 to record w eekly traffic volum es. Once each week cameras were checked and film was replaced as necessary* Water levels within each culvert if ere also recorded at this time as was the information collected on the traffic counter and the data loggers* the latter of which were reprogrammed at this time for the following week* The species of animal captured on each slide, as well as the date and time of each event, was recorded for each culvert using a stereomicroscope* and unique images w ere scanned in to digital files* Spreadsheets of environmental data were also constructed so that a comparison could be made between animal activity and these environmental parameters* An assessment of small-to- medium sized mammal species present adjacent to each main culvert pair* and a rough index of individual species abundance of the smaller species, was periodically made using live-trap transect lines and selective use of larger wire mesh traps* Twenty-five Sherman*' live traps (H, &. Sherman Traps, Inc*, Tallahassee, FL) were placed at 5 m spacing adjacent to and centered at each culvert entrance, baited with rolled oats and checked morning and evening for 4 days per session (Fig. 4), Each trap checked (morning or evening) constituted 1 "trapping opportunity"* Animals captured were identified to species* weighed, sexed, and aged (juvenile, subadult, adult) before release at the point of capture (no marking of individuals was made and thus no mark/recapture population estimates were attempted)* Tomahawk*' traps (Tomahawk Live Trap Company, Tomahawk, Wisconsin)/ baited with sardines, if ere also periodically set at culvert entrances to document presence of medium-sized mammals (e.g. , weasels/squirrels/skunks)* Figure 2- (See following page) Location of culverts in which cameras were installed. Culverts containing shelves and culvert dimensions are identified in the figure legem belli tf. - 0*9 m diameter culverts (and controls)* **' 1*2 m diameter culverts with existing shelves HI* 1*3 m to 2-2 m diameter culverts. - culverts selected for shelf additions. -fil - 2*3 m to 3*75 m diameter culverts. • .-, .. . - 3iu ^J 'r- a?v * ■•. r -^ U- S -'<• £ ;.. ,■:... • - ' } - 'I -A rZ y\ - '. Table 1. Specifications of all culverts studied. Site GPS location ft Lanes Shelf Substrate Length Width Height fess Creek Large (BCL) T J46*34 902 Wl 1 4*05.353 2 Yes None 27.6 m 3.7 m 2.3 m 'lore nee Bridge Culverl (FLO) 446*37,918 Wl 14*03.593 2 No Dirt (?of lenglh) 25.0 m 2.1 m i .0 m (i .5 m w/oul dirl) ^arleionCrk. Experimental kccEj 446*40.714 Wl 14*04.408 4 Yes None 59-5 m (63.2 m stelllength 1.2 m 1.2 m Carle ion Crk. ConTol kcccj 446*41 .01 5 Wl 14*04.389 4 No None 60.5 m 0.9 m 0.9 m Gravel Pit Control (GPC) 446*4t .273 Wl 14*04.369 4 No None 48.6 m 0.9 m 0.9 m 3ravel Pil Experimenial IGPE) 446*41 .472 Wl 14*04.355 4 Yes None 49.2 m (53.1 m Shell length 1.2 m 1.2 m Bravel Pil Large Culverl GPL) 446*41 .51 2 Wl 14*04.352 4 Yea None 50.0 m 1.4 m 15 m Aaclay Rat Experimental (MCE) 30 Large Culverl (I90L) 446*43.563 Wl 14*04.644 N46*57 .769 Wl 14*09.039 4 4 Yes No None Drl 50.0 m 65.0 m 1.2 m 3.5 m 1.2 m 3.75 m bo Small Culverl (I90S) 446*56.627 Wl 14*07 .053 4 No Dirt and Stones (to o! length) 79.0 m 0.9 m 0.9 m touHe Culvert i (DC1) J46*56.276 Wl 1 4*05.924 4 No Dirl and Stones 61.5 m 2.2 m 0.95 m Double Culvert 2 (DC2) 446*56.272 Wl 14*05.925 4 No Dirl and Stones 61.5 m 2.2 m 1.3 m The shell in BC L was removed each May and June to meet 50-year lood pJans Figure 3* Remotely activated cameras mounted in the culverts: 35 mm (a), and video (b). Figure 4* Small mammal live -trapping design: transect trap lines illustrated by dotted lines. Objective #2: Development and test of effectiveness of small mammal tubes to accommodate selective species* Preliminary research (40) indicated that at least 1 small mammal species abundant adjacent to the culverts* the meadow vole (Njcrotos p&nnsytvanicos), would not use the culverts to move under the highway. This avoidance behavior was felt to be due to this species' preference for protective cover so passageways were designed to provide a safe" environment through which these animals could travel* Such "vole tubes" were initially constructed from 3 m sections of plastic rain gutter dow nspouts connected together and suspended from the undersurface of the shelves (Fig. 5a,b)* These tubes were accessed by 1*2 m wide :: 0*05 m high plywood funnels w hich narrowed at the point at which they connected to the tube (Fig* 5c)* Funnels extended in to the surrounding vegetation allowing voles to enter without leaving protective cover* Animal use of these tubes was assessed by a carbon-sooted tracking plate Inserted in the middle section (approximately 25 m from each entrance; Fig* 6a, b); tracking paper was checked and replaced each week* Figure 5* Prototype design of vole tubes and entrance ramps: (a) gutter tubes and (b) connected tubes suspended from shelf and (c) entrance ramp to vole tube* ■m Ml* ' Figure 6* Sooted -tracking plate (a) placed in center section of prototype vole tube (b). Objective #3: Determination tfthe importance of vegetative cover at, and adjacent fo» culvert entrances* Vegetative cover measurements were made at entrances to main culverts 3 times over the course of the study. A visual estimate of percent cover » as assessed in quadrants at 0-3 m and 3-6 m distances from the entrances (Fig- 7)* Height of vegetation was also recorded by meter stick. Figure 7* Vegetation quadrants sampled at entrances to each main culvert. RESULTS Animal us* of culverts: Over the duration of the study more than 4,531 photographs and 8 r 135 total events were obtained of 14 different small mammal species (and 1 reptile, painted turtle - Chrysomys pica) using the main culverts. When all 12 culverts are considered, 7.510 photographs and 14/043 total events were recorded of 24 different species (Tables 2 and 3)* The additional species observed in the larger suite of culverts in part reflects the culvert's larger sizes and thus use by larger species (e.g.. deer bear)* Table 2. Summary of animal activity incillvertf from October 2001 through January 2004. :cc CCE GPC GPE 2PL MCE 90L 90S XM DC 2 BASS FLO *of Photos with Animals 872 773 679 724 1117 896 508 536 560 337 463 668 Total # of Photos 1374 1580 1015 1454 1752 2023 964 1321 1083 653 1026 1090 Estimated Total Activity 1 1511 1738 1117 1599 1927 2225 1060 1453 1191 718 1129 1199 T Estimated total activity equals the total number or photographs taken (with or without identifiable animals) + the average number of additional events recorded monthly by detectors. The latter value was determined by taking the average % or additional detections recorded over a 6 mo. period and then estimating this percentage Tor the additional 17 mo. and adding this value to each respective month. A definite seasonality in culvert use ™ as observed, with an increasing number of events occurring during the late winter and spring (February through May; Fig* 8)* The noticeable dips >r hich occurred in June 2002 and April-to-June 2003 can be explained by increasing water levels. With increasing spring snow melt and spring and summer rains, ground water levels rose in the wetlands adjacent to the main culverts* and the creeks at other culvert locations rose/ affecting animal use (Fig* 9). Culverts without animal shelves became impassable for smaller species. Activity which occurred when water levels were high (e.g., March and June 2002) reflected either use of shelves in the experimental culverts to walk above the vt ater (their intended purpose) and/or use of a control culvert without water at the same time that there was no use of one carrying water (i.e., the data is expressed as an average across all main culverts)* Deer mice, raccoons, striped skunks, short-tailed weasels, and domestic cats were the dominant species using the main culverts; in addition, coyotes* red foxes, and white-tailed deer commonly used larger culverts (Table 3)* Table 3. Animal us* of culverts as documented by photographs (October 2001 through January 2004), Culvert 1 . Species* PEMA J MIPE 3 PRLO' FEDCT WLUER 3 MEME 1 TAHLT ON2l ] GALA CAFA vuvu ODSP ME 185 18 71 29 216 24 O 348 GPL 952 24 O 24 55 18 O 2 o O GPE 630 6 3 12 51 io O GPC S44 31 25 SO 10 3 o o o O ccc SOO 10 30 118 2 137 2 5 CCE 172 1 ISO 232 12 156 a O O BOL 6 33 89 O 131 u 6 56 35 31 124 bos 392 7 2 18 O 59 O 33 1 19 O OCl 401 35 20 39 22 20 2 O 6 o DC2 97 1 92 93 2 32 1 2 o 13 o FLO 106 2 106 198 O 219 14 c 2 1 BCL 5 205 8 7 7 165 19 2 33 Total 4010 135 777 925 367 823 186 385 89 31 70 157 Culvert 1 . Species' SPCO : SORX : MAFL SYNU 3 MUVI ] NECI ] ERDO : TAAM S1NI* URAM MAPE CHPI 5 ME 1 3 O O O O O GPL 3 3 O 7 O 8 O O O O GPE o o o 6 o o o o c 6 GPC A 2 o o < c o o CCC 66 O o c o CCE 4 1 2 1 o box O 5 1 o bos 5 O o OCl o o G 9 o 4 1 o o i o DC 2 c c o 2 1 i c c o o FLO 7 9 4 o BCL O 5 2 2 Total 87 14 11 23 11 7 13 1 2 1 2 6 'Culvert notations as per Table 1, page 6. species notations: PEMA (deer mouse). MIPE (meadow vote), PRLO (raccoon), FEDO (domestic cat). MUER (snort-tailed weasel), MEME (Wiped skunk), TAHU (red squirrel), 0N2I (rnuskrat), CALA (coyote), CAFA (domestic dog), VUVU (red fox), OOSP (white -tailed deer), SPCO (Columbian ground squirrel), SORX (Shrew sp.), MAFL (yellow -Del lied marmot), SYNU (mountain cottontail), MLFVl (mink), NECI (bushy-tailed woodrat), ER0O (porcupine), TAAM (yellow pine chipmunk), SINE (fox squirrel), URAM (black bear), MAPE (fisher), CHPI (painted turtle). species identified using animal shelves. .'i 3 U J 1 4 ,rnnii*aT«n*c*^T -ar a in TS — — 1 1 \ < C 1 Q i A * •*i a **- o i A Q □ Q i ir D o ft H * n 1 'n \ A □ D cr T D ► i A D -4 D * LD \ \ V* / * V -\ -U- y -u- °\ •* /////////////•////////////// Figure 8* Seasonal activity pattern of animal use for all culverts, [(l) = average of total number of animals identified by photograph; (2) = average of total number of photographs with or without identifiable animals; Estimated to&f activity = (2) + average number of additional events recorded monthly by detectors. [The latter value w as determined by taking the average % of additional detections recorded over a 6 mo. period and then estimating this percentage for the additional 17 mo. and adding this value to each respective month], Figure 9. Animal activity in the culverts as a function of standing w ater levels (using only culverts CCC CCE GPG GPE GPQ. As would be expected* during warmer months animals were active over a xider range of temperatures and their activity was not concentrated around the monthly mean temperature (within ± 1 S*E») to the same degree as it was during colder months (e.g. compare July 2002' 2003 to February 2002/2003. Fig. 10 a-d). [refer to Figure 10 legend on next page] ,: * J SO- 1 ■ s " DO n n 1 r ■ J Hi m 1. grL- rrP llfn B *n <» i a ID V JB Jl 19 »!■■■!■■ 31 (O *> 10 » 44 BT a »' D II ■> T* *B« »^P Figure 10* Activity as a function of environmental temperatures in February (a) and July (b) 2002 and February (c) and July (d) 2003* [Vertically crossmatched bars = mean temperature: diagonally crossmatched bars = ±1 standard deviation]. O: course the primary question is whether or not animal use of culverts is enhanced by the presence of shelving* In all cases, as Mater levels rose in the control culverts animal activity decreased. Some species (e.g. raccoons) will still wade through these culverts >i rth u small amounts of water present (5-10 cm; Fig* 11) but even this small amount of water will preclude such culvert use by the smaller species (e.g.. deer mice and short-tailed weasels)* Activity in the experimental culverts however remained high or even increased when water levels rose, due to consistent use of the shelving* Atotal of 15 different mammal species were documented using the shelving (Table 3). Figure 11* Raccoon wading through Gravel Pit Control culvert this amount of water would be prohibitive to smaller species* Still photos and video sequences provided insight in to how smaller species were using the shelving* From this information, it became obvious that changes in the floor mesh were necessary to accommodate the smallest spedes* The 25 mm diamond mesh proved to be too open to allow rapid, easy movement of mammals the size of mice and weasels and these species were forced to move along the solid frame (Fig* 12 a.b). The solid plastic surface solved this problem and was readily used by all species (Fig- 13 a.b) but was not felt to be a permanent solution since it was not permeable to water if levels rose above the shelf. The 6 mm diamond mesh provided an appropriate surface for the smaller species but as used, was not a heavy enough gauge to support the larger species. Ultimately a # 13 flat galvanized expanded metal mesh was chosen which was accepted well by all species (Fig* 14 a.b)* ill Figure 12* Examples of 2 species using original shelves: Short-tailed weasel (a), and dc mouse (b), using the frame rather than the floor mesh of the original shelves* .4 Figure 13* Examples of 2 species using solid plastic surface: Deer mouse (a), and painted turtle (b). Figure 14* Examples of 2 species using modified shelves: Mountain cottontail (a)* and porcupine (b)* Shelf entrances were originally positioned directly in line with the shelf proper by MDT (Fig. 1 b) but it quickly became obvious that these would slow water flow* trap debris, and make shelf access impossible when water levels rose* Entrances were repositioned to the side of the culvert so that animals moving along the barrow pit could walk up the ramp and on to the shelf (Fig, 15)* From these results, a final design for the shelving was developed (Fig* 16; Appendix III)* T>r o of these net-; shelves were purchased from Roscoe Steel & Culvert Co*. Missoula- MT and placed in the GPL and &CL culverts for further assessment* Consistent continued use of these shelves by a variety of species has been documented by still and video cameras (e*g.* Fig. 14 a.b). For a short period of time (approximately 3 weeks) the entrance ramps to the GPL shelf were unavailable and yet animals still climbed up 0*6 m to use the shelf when water was present. Figure 15* Entrance ramp at Carleton Creek Experimental positioned to side of shelving and culvert* 15 Figure 16* Ftiotos of new shelf and entrance: (a) frontal view with entrance ramp entering from side, (b) side view, and (c) top view showing width of entrance ramp and shelf. Effective ne ss of small mammal shelf tubts "Vole tubes" accessed by an entrance funnel extending in to the vegetation at culvert entrances were heavily used by meadow voles deer mice, and short-tailed weasels (Figs. 17 and 18)* Hundreds of tracks were obtained on the tracking paper each week* These tubes worked so well in overcoming the avoidance behavior of voles that they were incorporated as an integral component of the frame used in the final shelf design (Fig* 19 a.b). Figure 17* appearance of entrance funnel entering from side (plywood model in middle of photo)* [Compare this prototype to the final design illustrated in Fig. 18b where the funnel is incorporated in to the entrance ramp] US I ^ _ Figure IS* Animal use of vole tubes illustrated by sooted-tracking plates; (top) meadow vole and deer mouse tracks, and (bottom) short-tailed v easel tracks and drag mark of prey* Figure 19* Ooss section (a) of final shelf design illustrating vole tube incorporated in to the shelf frame and access funnel (b) built in to the shelf s entrance ramp* r Animal activity at additional culverts The 7 additional culverts added during the course of this research spanned a >tide range of sizes and configurations (Table 1)* Comparison of animal activity between these provides considerable preliminary information since some Mere dry culverts with gravel or sand floors, and others serviced permanent streams and were large enough to accommodate species such as deer. All of these culverts were actively used by a Hide variety of animal species (Tables 2 and 3). Two results in particular stand out. First, can ids. such as red foxes and coyotes, routinely used those culverts which had a more natural ground cover [e.g.. I90L I90S and FLO; Fig. 20 a.b)* Second, >i hite-tailed deer routinely used 2 of the largest culverts I90L and BCL (Fig- 21 a . b ) ; interestingly the former has a deep layer of silt built up covering the corrugated steel floor while the latter* servicing Bass Creek, is continually scoured such that the corrugated floor has remained clean* At BCL deer walked in the stream through the culvert on the corrugated metal surface* Figure 20* Red fox (a), and coyote (b) using 190 Large culvert v rth a silted floor surface. Figure 21* White -tailed deer using 190 Large (a).- and Bass Creek Large (b) culverts* It if as not uncommon for numerous species to use a particular culvert on a nightly basis* each moving through at a different time* For example, white-tailed deer, raccoons* striped skunks, red foxes, coyotes, muskrats, domestic dogs and cats were all observed to use I90L over a short period of time. As indicated above* the final shelf design w as tested in 2 additional culverts (GPL and 8CL Fig* 31 a.b. Appendix II)* Both culverts w ere somewhat larger than those in the original study, BCL in particular was a 2.3 k 3*7 m steel corrugated squash culvert in which the shelf was hung approximately 1.5 m above the floor (Fig. 22). Within 1 day after placement this shelf was being used by numerous animals* In agreement with MDT* the shelf in BCL was removed during the May- June period to accommodate the 50-year flood plans for Bass Creek* Figure 22* Modified animal shelf placed in Bass Creek Large culvert Importance of vegetative cower at culvert entrances Percent vegetative cover (Fig. 23) and height of vegetation (Fig* 24) at the entrances to the main culverts originally varied greatly between sites* Vegetation on the eastern side of the highway was less disturbed during construction than that on the western side since a rail line lies along the eastern boundary and highway expansion subsequently had to occur in the westerly direction (see Figs* 2S-32, Appendix I)* Over the duration of the current study, following the reseeding efforts of MDT* vegetation on the western side began to take hold effectively* Vegetative growth, of course, varied seasonally thus measurements Mere only recorded during the spring through fall months* Percent cover increased as a function of distance from the culvert entrance (Fig. 24 a vs. b). A strong correlation was also observed between percent cover/height and overall animal activity (Fig. 25)* Those culverts exhibiting the greatest vegetative cover (GPE, CCE and GPL- adjacent to GPE but not included in the main culvert design) were also the ones which exhibited the greatest amount of animal activity. One exception was noted at the MCE, and MCE,., entrances - here, because the culvert it as originally placed too low in the roadbed* standing water was continually present* The only vegetation at this entrance was either off to the sides or represented by cattails projecting through the water* i*' icw ictt" tu oru am ora ar^i j^ ax* act i» ckc *** 1 I-. !- * > 1 ca^ i* opm art* arar Oft* flfC* arc* ceo* ccc* CCC* CCt> Figure 23* Vegetative cover at the entrances to each main culvert: (a) average % (of 3 surveys) of vegetative cover within 3 m of entrances and (b) average % (of 3 surveys) of vegetative cover within 3-6 m of entrances (culvert abbreviations as per Table 1, page 6) [*MC£, and MCE,, entrances v ere completely covered in water thus vegetative cover reflects dumps on the edges or cattails projecting through the water]. 20 «^ «:&■ jn* -a* u ■::■:» . :v . t. ittk iceb i"L* s p Lt arte a*a or^A ortt so* csct not rci Figure 24* Height of vegetation at the entrances to each main culvert: (a) height of vegetation within 3 m of entrances, and (b) height of vegetation within 3 to 6 m of entrances (culvert abbreviations as per Table 1* page 6)* (MCE., and MCE entrances >* ere completely covered in Mater thus vegetative cover reflects clumps on the edges or cattails projecting through the water]* :i *a-i Figure 25* Relationship between vegetative cover adjacent to entrances of the main culverts and animal activity (as indexed by total number of photographs recorded; culvert abbreviations as in Table 1, page 6). Traffic ro fumes The traffic counter worked effectively between June 5, 2002, when it was installed.- and December 18, 2002; at this point it began to give erroneous readings and was no longer employed' Weekly traffic counts however were very consistent during the summer and fall months averaging approximately 6.000 vehicles per day (both directions were recorded)/ with a peak occurring in mid-July (Fig* 26)* Traffic levels were higher in early summer and noticeably declined in early winter. Since there were no exit points from the high ii ay betu een the main culverts these traffic volumes accurately reflect vehicle loads along this highway section* No attempt was made to record hourly volumes though there are obvious peak periods between 6:00 - 9:00 am and 4:00 - 7:00 pm as travel from the Bitterroot valley to and from Missoula occurs. However* nearly 98% of the animal activity through the culverts occurred during darkness, generally between 9:00 pm and 3:00 am, a reflection of the nocturnal behavior of most species. Animals may thus be automatically avoiding these heavier traffic periods* Ms* nira /ilj. . Species Relative Abundance Meadoir voles 1328 Deer mice 742 Short-tailed weasels 26 Vagrant shrews (Sorex vagrtns) 81 Columbian ground squirrels 38 Western lumpinq mice (Zapusprsnceps) 1 House mice (IHjs mucsulus) 1 :; 1 1 ! o V . «^- B 30V- *]\ ^ V] ^ ■SFCO OSORX - OyuER OFEUA QWIPE ^^ 1 ^^r 1 1 1' Figure 27. Summary of small mammal live -crapping sessions over the period March 2001 to September 2003, [Animal notations as per Table 3/ page 10] CONCLUSIONS High h ays by their very nature fragment the habitat through w hich they are built. Numerous studies have shown that such fragmentation lessens permeability to animal movement and may lead to a decrease in species diversity imparted by demographic changes as well as longerterm genetic effects (3,9,10,41,42,43,44,45,46,47). The present research was designed to address and hopefully mitigate for such fragmentation, specifically for small mammal populations. Specifically, the effectiveness of ramp/shelf structures within drainage culverts in allowing the movement of small mammals under a 4-lane highway during periods of water flow if as studied* Initial research suggested several shelf modifications needed to be made; once these were addressed this design proved highly effective! allowing a wide variety of small mammals inhabiting the wetlands adjacent to the highway to move from one side to the other safely (Table 3; Figs* 12-14)* Over 8,000 events (movements past detectors) were recorded representing 14 different mammal, and 1 reptilian species* Movement patterns were highly seasonal (Fig- 8)* Multiple factors may be responsible for these observed patterns* Decreased activity in late fall/early winter (November* January periods) most likely reflects periods of colder temperatures* Three species* the Columbian ground squirrel, yellow -pine chipmunk* and yellow-bellied marmot, hibernate during the fall and winter months and are thus not active during this period; others such as the deer mouse (the second most common species in the wetlands and the most common to use the culverts) will reduce its activity under colder conditions, entering nightly torpor, while striped skunks* porcupines* and raccoons will reduce their nighttime foraging as well (43)* This is borne out by the observation that most activity during colder periods was concentrated within a narrow range of temperatures: in contrast, warmer :4 temperatures provided a much wider range over which movement would occur (Fig. 10)* Greater activity was observed in late winter and early spring (February-April 2002, February 2003) possibly reflecting the onset of breeding for many species and thus increased movement of males. Warmer temperatures, and little snow, w ere recorded during these months (the exception noted for March-April 2003 reflects significant elevation in water levels). Decreased activity in June and July may be due to lessened movement by pregnant/post-parturient females and rising water levels in and adjacent to the culverts due to spring snowmelt and rainfall (particularly the control culverts without shelving; Fig. 9)* On the east side of the highway the barrow pit is narrow (** 30 m in width) sandwiched between the highway bed and the railroad line* When the water table rises in the wetlands, this area is often under 0*5 -1 m of standing water restricting access to both control and experimental culverts. Culverts with shelves continued to record a high amount of activity, especially when water levels rose (Fig* 9). Initial observations on how the smallest mammals walked on the shelving suggested that the 25 mm diamond -shaped* extruded metal shelf surface was too open for easy movement* Tests wrth a solid surface confirmed this observation and led to the test of a 6 mm mesh size. In order to meet strength constraints, the final material chosen for the shelf surface if as a * 13 flat galvanized expanded metal mesh* In addition, it was determined that entrance ramps had to be moved to the side of the culverts to allow unimpeded water flow* and access to the shelves when water if as present- This design proved to be very effective (Figs* 15 and 16): video sequences of animal movement following these modifications further supported its effectiveness* The most common species inhabiting the wetlands adjacent to the culverts if as the meadow vole yet this species was only observed in dry culverts on a few occasions. Development of vole tubes which allowed movement from the wetlands into and through the culvert without leaving protective cover also proved to be highly effective (Figs* 5, 15, 16, and 19)* The final shelf design incorporates these vole tubes in to the frame and the entrance funnel in to the surface ramp used by larger species* The entrance ramp is hinged so that it can conform to each site's topography. Built in 2*5 m sections, the shelving is designed to be easily inserted in to existing culverts so that they can be retrofitted for animal use (Appendix III). The characteristics of larger, steel corrugated culverts were also studied in an attempt to determine their effectiveness for animal movement Larger animal species seemed, in general, to behaviorally avoid such culverts* most likely because of the difficulty in walking on the hard* slick, corrugated surface* However* those culverts which had built up a layer of silt due to slow or intermittent water flow (notably I90L and FLO) were used routinely by a large number of species* 190 Large is particularly interesting because it lies under a 4-lane divided section of highway, and is fully 65 m in length. A total of 11 different species routinely used this culvert among which was white-tailed deer* This finding supports results of studies conducted in Banff National Park* Alberta which analyzed ungulate use of dry culverts [38)* It must be noted however that in the cited study animal movement was dictated by continuous fencing which was not the case in our studies* In the current study* there was no indication that species such as deer respond to the "tunnel effect" (the visual impression that the distant opening of the culvert is too small to exit) created by the length of this culvert (Fig* 21a)* Canids* as well, readily used these culverts on a daily basis (Fig* 20 a*b); one coyote so consistently used the I90L culvert that it would even swim through it when the creek it served (CKeefe Creek) rose in the spring* Another large culvert* BCL is also of interest as this one serves South Bass Creek, and in so doing has continuous water flow* A majority of time there is 15-25 cm of water flowing through BCL increasing to 1+ m during spring runoff* This volume is adequate enough to prevent any sediment from building up between the corrugations yet deer were often seen moving through (Fig* 21 b)* This behavior further suggests that no "tunnel effect" is perceived and that deer will actively use such culverts. Most likely the riparian corridor along the southern tributary of Bass Creek, which is heavily wooded on both sides of this culvert di reedy up to both entrances/ serves as a natural, protective funnel for deer (and other spedes as well e.g. - raccoons fisher)* This illustrates the importance of adjacent vegetation to animal use of culverts. Most wild species use vegetative cover to either hide from predators or sneak up on prey* As such, they avoid open areas as are often found at the entrances to culverts* Earlier studies identified vegetative cover near tunnels/culverts as one of if not the most important attribute influencing small mammal use (25,77,33,49)* Our studies support this finding. demonstrating a direct correlation between use of the main culverts and the % of vegetative cover present (Fig* 25)* One interesting observation was made between meadow voles and deer mice, the predominant species in the wetlands. These two species exhibit nearly completely opposite behaviors: meadow voles hide under vegetation and move along runw ays w hich they themselves create* totally avoiding open areas and thus culverts (10 ,4$); deer mice* however, prefer disturbed and open habitats and* as such* freely move across open areas at the entrances to culverts and enter them as has been documented (18,40,48)* Vole tubes* as described above, solved this avoidance behavior of meadow voles. The volume of traffic along a roadway translates in to a variety of disturbances such as increased noise levels, increased pollution* and a general increase in overall visual disturbance (24,43)* These factors have been shown to affect animal populations in adjacent areas creating, in many cases, what is known as a "road avoidance zone" (24,43,50,5 1,52,53,54)* Some studies suggest that animals may become habituated to a constant predictable level of disturbance (55,56,57) but this is not borne out by other studies (54)* Traffic volumes along the section of Highway 93 served by the main culverts in the current study remained relatively constant* though a slight elevation » as observed in early summer (June/July), while a slight reduction occurred in early winter (November/December) (Fig* 26)* No correlation between this activity and animal use of these culverts was apparent (compare Fig* 26 with Fig* 8)* Though this section of highway was expanded from 2-lanes to 4, and would be expected to carry an increased traffic load in the future< it is unlikely that increased volumes occurred just during, or immediately following this expansion, the time period encompassing the course of this study, It would be interesting to revisit this issue along this section of highway in 10-15 years as increased development in the Bitterroot Valley, and thus increased traffic volumes* continues. Live-trapping of small mammal populations adjacent to the main culverts over the duration of this study provided the data necessary for determining which species would effectively use these structures to circumnavigate the highway. As previously explained, this information was useful in determining the need for* and the creation of, tubes for meadow voles. Small mammal use appeared to be directly proportional to the relative abundance of the species (Table 4) with one exception, that of the Columbian ground squirrel. This species was trapped periodically during the summer months, specifically adjacent to the west entrances of GPE and GPC Transect lines placed here lay directly adjacent to drier* cultivated fields w hich supported a large population of this species* The habitat on the eastern side of these culverts was much more moist, and more heavily vegetated, inappropriate habitat for this species* and thus litde movement in this direction occurred. The trend which was noticed in deer mouse and meadow vole populations from spring through fall (declines in deer mice/increases in meadow voles) may be explained by the different reproductive characteristics of these species* Deer mice in Montana begin breeding in March and may produce 3-4 litters over the course of the spring and summer (48)* Meadow voles however can breed throughoutthe year and they exhibit a shorter gestation period and larger litter sizes* The trend in relative abundance may have also been influenced by mowing of the vegetation by MDT immediately adjacent to the pavement. This area was strictly inhabited by deer mice while meadow voles were restricted to the more mesic, lower sites* The paucity of Columbian ground squirrels in the spring and fall Zfi compared to their abundance during the summer simply reflects the fact that they hibernate from early August until mid-April {43). RECOMMENDATIONS Three recommendations clearly arise from the results of this study: (1) Small mammal shelving with built-in vole tubes should be employed in all culverts where habitat characteristics are appropriate and preclude the installation of dry culverts (eg., wetlands adjacent to the highway, permanent water sources which are serviced by a culvert). The cost of such structures is minimal and their effectiveness in allowing animal movement under the highway is proven. Steel corrugated culverts can easily be retrofitted with such shelving. If placed in larger culverts, small mammal movement can be accommodated along with the movement of larger species (see recommendation #2 below) at a fraction of the cost of bridge -type wild/ire structures. (2) Evidence suggests that species such as deer will readily use large, steel corrugated culverts if these are tied to protective vegetation at their entrances. Further studies should be conducted to determine the best way to modify the floor corrugations to provide a more natural surface. Once appropriate modifications are identified, it is felt that this type of wildlife structure would be as functional as bridge-type structures for a fraction of the cost. (3) Vegetative cover at culvert entrances is required in order for most, if not all, species to access them. Additional studies would be useful to identify the most appropriate plant species that would provide such cover while not impeding water flow. IMPLEMENTATION Small mammal shelves, as described here, have now been incorporated in to the final engineering plans of Peccia & Associates for Phase 5 of the High ir ay 93 South reconstruction from Florence to the Stevensville Wye. Additional such shelves have also been identified for placement along the Conner section of highway south of Hamilton, Montana. ZT REFERENCES CITED (1) Simmons* J* R*. Feathers and fur on the turnpike. Christopher Publishing House* Boston* (1938) (2) Knobloch, W., "Death on the highway." Journal of Mammalogy, Vol* 20 (1939) pp* 508- 509. (3) Haugan, A* O*, "Highway mortality of wildlife in southern Michigan.* Journal of Mammalogy, Vol* 25 (1944) pp* 141-146. (4) Hod son, M- Li "A survey of road mortality in mammals (and including data for the grass snake and common frog)*" Journal of Zoology, London* Vol* 148 (1966) pp* 576-579* (5) Bellis, E* D*, and Graves, H* 8*.- n Deer mortality on a Pennsylvania interstate highway*" Journal of Wildlife Management, Vol. 35 (1971) pp* 232-237* (6) Conover, M. R*, Pitt, W. G. Kessler, K* K*. DuBow, T. J., and Sanborn. W. A** "Review of human injuries/ illnesses and economic losses caused by wildlife in the US." Wildlife Society Bulletin, Vol 23 (1995) pp. 407-414* (7) Hughes, W* E*, Saremi, A R** and Paniati, J* F„ 1996* 'Vehicle-animal crashes: an increasing safety problem*" Institute of Transportation Engineers Journal, Vol. 66 (1996) pp* 24-28* (8) Paqueti P.* and Callaghan, C, 'Effects of linear development on winter movements of gray wolves in the Bow River Valley of Banff National Park* Alberta." Proceedings of the Transportation Related Wildlife Mortality Seminar, State of Florida Department of Transportation! Environmental Management Office, Tallahassee, Florida* FL-ER-58- 96. (1996) No pagination* (9) Mattson. D- Knight J. R* and Blanchard, B* "The effects of developments and primary road systems on grizzly bear habitat use in Yellowstone National Park* Wyoming*" International Conference on Bear Research and Management, Vol* 7 (1987) pp* 259- 273. (10) Oxley* D* J*, Fenton, M* B*, and Carmody, G. R. r v The effects of roads on populations of small mammal" Journal of Applied Ecology, Vol* 11 (1973) pp. 51-59. (11) Werner* E. F*, "Mammals of the thousand island region. New York*™ Journal of Mammalogy, Vol* 37 (1956) pp* 395-406. (12) Sheppe. W., "Island populations and gene flow in the deer mouse* Peromyscus leucopus" Evolution, Vol 19 (1965) pp* 480-495. (13) Brehm, K., "The acceptance of 0.2 metre tunnels by amphibians duri to the breeding site** Proceedings of the Toad Tunnel Conference, jring their migration e, ACO Polymer Products, Bedfordshire, (1989) pp* 29-42* (14) Camby, A*, and Maizeret, G, "Permeabilite des routes et autoroutes vis-a-vis des mammiferes carnivores: Exemple des etudes menees dans les Landes e Gascogne." Routes etfaune sauvage. Actes du collogue* Minis tere de ("Equipement, du Logement du I'Amenagement du Territoire et des Transports* Bagneux-Cedex. (1985) pp* 183-203* (15) Hunt/ A, Dickens, H. J,, and Whelan, R. X, "Movement of mammals through tunnels under railway lines." Australian Zoology > Vol* 24 (1987) 89-93* (16) Podloucky R* "Protection of amphibians on roads: examples and experiences from Lower Saxony.* Proceedings of the Toad Tunnel Conference, ACO Polymer Products, Bedfordshire* (1989) pp. 15-28* (17) Waechter, A* and Schirmer, R* r "Mortalite animate et effects de coupure (differents types de mortality insularisation de la faune, strategies adaptatives)*" Cfcseivatoire ecologique de la Basse Vallee de la Doller* Actes du collogue* Minister* de I'Equipement, du Logement, de Transports et de la Mer, Metz* (1987) pp* 100-116* (18) Yanes, M., Velasco, J* M.. and Sua res. F* "Permeability of roads and railways to vertebrates: the importance of culverts* 1 * Biological Conservation, 71 (1995) pp* 217- 222 (19) Evink, G. L, Zieqler. D*, Garrett P. and Berry. J„ Proceedings of the Florid* Department of Transporta tion/ Federal Highway Administration Transportation- Related Wildlife Mortality Seminar* FL-ER-58-96. Florida Department of Transportation/ Tallahassee, Florida* (1996)* Mo pagination. (20) Evink, G. L, Garrett P*, Ziegler, D., and Berry, R., Proceedings of the Florida Department of Transportation/ Federal Highway Administration Transportation- Re la ted Wildlife Mortality Seminar. FL-ER-69-98* Florida Department of Transportation. Tallahassee, Florida, (1998). No pagination (21) Evink, G* L, Garrett P.. and Zeigler, D.. Proceedings of the Third International Conference on Wildlife Ecology and Transportation* FL-ER- 73-99. Florida Department of Transportation, Tallahassee, Florida, (1999) pp* 329* (22) Evink, G> L Proceedings of the International Conference on Ecology and Transportation, Keystone, CO., September 24-28, 2001; Raleigh, MC Center for Transportation and the Environment NC State University. (March 2002)* pp. 675* (23) Evink, G* L Proceedings of the International Conference on Ecology and Transportation, Lake Placid, NY, August 24-29, 2003* (In Press)* (24) Form an, R* T* T* et al* Road Ecology: Science and solutions* Island Press. Washington, DC. (2003) pp* 481. (25) Waller, J*, and Servheen, C "Documenting grizzly bear highway crossing patterns using GPS technology.* Proceedings of the Third International Conference on Wildlife Ecology and Transportation* FL-ER- 73-99* Florida Department of Transportation. Tallahassee, Florida, (1999) pp* 21-24* (26) Singleton, P., and Lehmkuhl, J* "Assessing nil d life habitat connectivity in the Interstate 90 Snoqualmie Pass corridor. Washington*" Proceedings of the Third International Conference on Wildlife Ecology and Transportation. FL-ER- 73-99. Florida Department of Transportation, Tallahassee* Florida, (1999) pp* 75-84, (27) Callaghan, C, Paquet* P.. and Wierzchowski, J*, "Highway effects on gray wolves Mithin the Golden Canyon, British Columbia*" Proceedings of the Third International 29 Conference en Wildlife Ecology and Transportation. FL-ER- 73-99. Florida Department of Transportation, Tallahassee, Florida (1999) pp. 39-52 (28) ChruszcZi B. T Clevenger, A P.. Gunson. K* £., and Gibeau, M. L, ''Relationships among grizzly bears* highways, and habitat in the Banff Valley. Alberta, Canada. * Canadian Journal of Zoohgy, Vol 81 (2003) pp. 1378-1391. (29) Land, D*. and Lotz, M. "Wildlife crossing designs and use by Florida panthers and other wildlife in southwest Florida*" Trends in addressing transportation related wildlife mortality* Proceedings of the Transportation Related Wildlife PlortaHty Seminar, State of Florida Department of Transportation, Environmental Management Office, Tallahassee, Florida* FL-ER-5S-96, (1996) Mo pagination. (30) Madsen r A* B* "Otter Lotra futra mortality in relation to traffic and experience with newly established fauna passages at existing road bridges** Lotra, Vol* 39 (1996) pp*76-8B. (31) Bonds, B*i "Wildlife habitat mitigation." Proceedings of the Third International Conference on Wildlife Ecology and Transportation. FL-ER- 73-99. Florida Depaitment of Transportation, Tallahassee, Florida, (1999) pp. 173-174* (32) Philcox, C K*, Grogan, A. L, and MacDonald, D, W.. 1999* "Patterns of otter Lutra futra road mortality in Britain.* Journal of Applied Ecology, Vol, 36 (1999) pp* 748- 762 (33) Wells, P.. Woods. J., Bridgewater, G*, and Morrison, H, "Wildlife mortalities on railways: Monitoring methods and mitigation strategies*" Proceedings of the Third International Conference on Wildlife Ecology and Transportation. FL-ER- 73-99. Florida Depaitment of Transportation, Tallahassee/ Florida, (1999) pp* 85-88* (34) Ruby, D* E.-, Spotila, J* R*, Martin, S* K., and Kemp, S* J*. "Behavioral responses to barriers by deseit toitoises: Implications for wildlife management** Herpetological Monographs, Vol. 8 (1994) pp* 144-160. (35) Boarman, W* L, "Highway mortality in deseit tortoises and small vertebrates: Success of barrier fences and culverts*" Trends in addressing transportation related wildlife mortality* Proceedings of the Transportation Related Wildlife Mortality Seminar t State of Florida Department of Transportation, Environmental Management Office, Tallahassee, Florida* FL-ER-58-96 (1996) No pagination. (36) Jackson, S* "Underpass systems for amphibians*" Trends in addressing transportation related wildlife mortality* Proceedings of the Transportation Related Wildlife PSortality Seminar, State of Florida Department of Transportation, Environmental Management Office* Tallahassee, Florida* FL-ER-58-96, (1996) Mo pagination* (37) Rodriguez, A, Crema, G* and Delibes, M. "Use of non-wildlife passages across a high speed railway by terrestrial veiteb rates*" Journal of Applied Ecology, Vol. 33 (1996) pp* 1527-1540. (38) Clevenger, A* P., and Waltho, N, "Dry drainage culvert use and design considerations for small- and medium-sized mammal movement across a major transportation corridor*" Proceedings of the Third International Conference on Wildlife Ecology and Transportation. FL-ER- 73-99* Florida Department of Transportation, Tallahassee. Florida, (1999) pp. 263-278* (39) Clevenger, A P., and Waltho, N. "Factors influencing the effectiveness of wildlife underpasses in Banff National Park, Alberta, Canada." Conservation Biology, Vol. 14 (2000) pp. 47-56. (40) Foresman.. K* R. "Monitoring animal use of modified drainage culverts on the Lolo South Project-" Final Report FHWA'MT-01-004/8117-15, Montana Department of Transportation and The U.S. Department of Transportation, Federal Highway Administration (2001) pp. 18. (41) Adams. L W.. and Gels A D. "Effects of roads on small mammals*" Journal of Applied Ecology, Vol. 20 (1983) pp* 403-415. (42) Garland, T., and Bradley, V, G* "Effects of a highway on Mojave Deseit rodent populations*" American Midland Naturalist, Vol* 111 (1984) pp. 47-56* (43) Form an. R* T* T, and Alexander, L E. "Roads and their major ecological effects/ Annual Review of Ecology and Systematics, Vol. 29(1998) pp. 207-231* (44) Mansergh, I. M*, and Scotts, D, J, "Habitat-continuity and social organization of the mountain pygmy possum restored by tunnel*" Journal of Wildlife tlanagement, Vol. 53 (1989) pp* 701-707* (45) Wilcox, B. A, and Murphy.. D* D* "Conservation strategy: the effects of fragmentation on extinction.* American Naturalist, Vol* 125 (1985) pp. 879-887* (46) Wood vi ard. S* "Population density and home range characteristics of woodchucks at expressway interchanges*" Canadian Field- Naturalist, Vol* 104 (1990) pp* 421-428. (47) Van der Zee, F* R, Wiertz, J*, Ter Braak, C* J* F*. van Apeldorn, R* G, and Vinlc. J, "Landscape change as possible cause of the badger /fetes mefes L decline in The Netherlands*" Biohgical Conservation, Vol* 61 (1992) pp. 17-22* (48) Foresman, K* R*, The Wild Mammals of Montana. Special Publication No* 12. The American Society of Mammalogists, Allen Press. Inc./ Lawrence, Kansas (2001) pp. 27a (49) Rodriguez, A, Crema, G*.- and Delibes, M* "Factors affecting crossing of red foxes and wildcats through non-wildfrfe passages across a high-speed rail if ay. Ecography, Vol* 20 (1997) pp* 287-294* (50) Rost G, R.. and Bailey, J* A "Distribution of mule deer and elk in relation to roads." Journal of Wildfire Management, Vol. 43 (1979) pp* 634-641. (51) Gese, E. M- Rongstad, O. X, and Mytton* IV* R. "Changes in coyote movements due to military activity," Journal of Wildlife Management, Vol* 53 (1989) pp. 334-339* (52) Get:, L L, Cote, F. R. and Gates, D* L "Interstate roadsides as dispersal routes for Microtus pennsyfvanicus" Journal of Mammalogy, Vol. 59(1978) pp. 208-213* (53) Mace, R* D*, Waller, J* S„ Manley, T* L Lyon, L* J*, and Zuuring, H. "Relationships .ii amoung grizzly bears* roads* and habitat in the Swan Mountains, Montana." Journal of Applied Ecology, Vol. 33(1996) pp. 1395-1404. (54) Gtbeau, M. L, Clevenger, A P., Herrero* 5.. and Wierzchowslci, J* "Grizzly bear response to human development and activities in the Box River Watershed* Alberta, Canada. Biological Conservation, Vol. 103 (2002) pp. 227-236. (55) Jope, K. L. "Implications of grizzly bear habituation to hikers*" Wildlife Society Bulletin, Vol* 13 (1985) pp* 32-37* (56) McLellan, B* N*, and Shackleton. D* M* "Immediate reactions of grizzly bears to human activities.* Wildlife Society Bulletin, Vol* 17 (1989) pp* 269-274* (57) Bowles, A* E* "Responses of wildlife to noise." Wildlife and Rea creationists: Coexistence through Management and Research, Washington. D. G, Island Press (1997) pp* 109-156. M APPENDIX I: Mai* Culverts in this study and vegetative cover attfieir entrances West entrance heavily vegetated East entrance heavily vegetated Figure 28* Carleton Creek Experimental Culvert West entrance unvegetated East entrance veil vegetated Figure 29* Carleton Creek Control Culvert West entrance heavily vegetated East entrance heavily vegetated Figure 30* Gravel Pit Experimental Culvert West entrance sparsely vegetated Ease entrance vegetated Figure 31* Gravel Pit Control Culvert West entrance heavily vegetated East entrance heavily vegetated Figure 32* Maclay Flats Experimental Culvert *' The Maclay Flats Control culvert nas placed too high in the roadbed to serve as an appropriate control as it is continually full of v;ater thus it nas omitted from this study* APPENDIX II; Additional culverts added during this study " i Rgure 33* Additional culverts along Highway 93 South: (a) Gravel Pit Urge, and (b) Bass Creek Large* Figure 34* Additional culvert added along Highway 203: Florence culvert R8 mk Figure 35* Additional culveits along Interstate 90: (a) 190 Large, (b) 190 Small and (c) Double culvert .17 APPENDIX HI: Final design of animal shelves. Figure 36* Final design of animal shelf: Individual 2.5 m sections were built vuth brackets to hang from the side of the culvert as » ell as from above (a)/ shelf in place illustrating connected sections as well as vole tube (b; see also Fig. 18a)/ roof bracket to support shelf and cable from ithich it is hung (c) side mounting bracket with pins to support shelf (d), and sheff in place (e). [Roscoe Steel 8c Culvert Co., Mssoula, MT; WMtf.roscoesteel.com]